This study used data from a large STEMI registry, a database that includes all consecutive patients admitted for STEMI at a regional referral heart center. This heart center is the leading interventional facility for MI patients in a region with a population of nearly 1 million. Initial data entry is performed by interventional cardiologists in the cardiac catheter laboratory and is subsequently supplemented by comprehensive data collection once a second cardiologist confirms the diagnosis.

Eligible for inclusion were all consecutive patients diagnosed with STEMI, admitted to the heart center between January 2015 and February 2022.

Demographics such as age, sex, and address-based data (Green Care Index, Social Deprivation Index [SDI]); clinical characteristics (body mass index [BMI], diabetes mellitus, hypertension, smoking status); lipid profiles (high-density-lipoprotein cholesterol cholesterol, low-density-lipoprotein cholesterol cholesterol, and total cholesterol); previous history of coronary artery disease; MI characteristics (Killip classification, number of diseased coronary arteries, and left ventricular ejection fraction [LVEF]); and interventional outcomes (postinterventional Thrombolysis In Myocardial Infarction [TIMI] flow and procedural success) were collected at the time of index catheterization. Concomitant diseases were recorded dichotomously as either present or absent.

SES was operationalized at the neighborhood level using the Bremen SDI, an area-based, city-wide composite assigned to each patient’s geocoded residential address. The SDI is officially calculated by the Senator for Social Affairs within the framework of the city’s Monitoring Program, which has a longstanding tradition of small-area social analysis in Bremen. Compared with other major German cities, Bremen provides unusually granular, small-scale data, allowing for higher-precision neighborhood classification. This approach is conceptually aligned with both the European Deprivation Index (EU–SILC-based) and the German Index of Socioeconomic Deprivation. ^,,,,

The SDI encompasses seven overarching dimensions of deprivation—income, employment, housing, education, political participation, safety/security, and social inclusion—each represented by standardized indicators reflecting the socioeconomic context of residence. Methodologically, the composite index is calculated as the mean of seven z-standardized indicators: need for language support in preschool screening, lack of an Abitur (upper secondary degree), safety/crime, SGB-II transfer density for individuals < 15 lt; 15 years, SGB-II transfer density for individuals ≥15 years, unemployment rate, and voter nonparticipation (political participation). Each patient’s residential address was geocoded to the corresponding city subdistrict and linked to its SDI value.

Following the official city-wide classification, SDI values were mapped to four predefined status categories using z-score cut-offs: high (z > +1), intermediate (−1 ≤ z ≤ +1), low (−1.5 < z < −1.0), and very low (z ≤ −1.5). For analyses and figure labeling, these categories were denoted as G1: High SES, G2: Intermediate SES, G3: Low SES, and G4: Very Low SES ( Figure 1 ).

Impact of socioeconomic status on long-term outcomes: adjusted survival analysis for cumulative incidence of major adverse cardiovascular events (MACCE). SES levels are color-coded: High SES ( red ), intermediate SES ( green ), low SES ( teal ), and very low SES ( purple ). The multivariate survival analysis for MACCE was adjusted for age, sex, traditional risk factors, infarction severity, and interventional success. A 90-day landmark sensitivity analysis showed consistent results (low SES HR 1.79, p = 0.004; very low SES HR 2.28, p = 0.002).

For readability and consistency with the cardiovascular literature, we use the umbrella term “socioeconomic status (SES)” throughout the manuscript. Accordingly, patients were categorized as G1–G4 as defined above.

Patients were followed for MACCE through telephone interviews and review of hospital readmission records, with follow-up extending to 60 months. Accordingly, the Kaplan–Meier curves depict event rates across the full 5-year horizon.

The primary outcome was the occurrence of MACCE during long-term follow-up, defined as a composite of MI, stroke, and all-cause mortality.

Data collection in the Bremen STEMI registry was approved by the ethics commission of the Bremen Medical Association in Germany. Study participants or their legal guardians provided written informed consent to participate in the study.

BMI was calculated as the quotient of weight (in kilograms) divided by height squared (in meters squared) for each patient.

A family history of coronary artery disease (CAD) was defined as coronary heart disease in first-degree relatives occurring before the age of 55 in males or 65 in females.

Late presenters were defined as patients with a pain-to-door time of ≥6 hours and/or indicative signs of subacute MI evident on the admission electrocardiogram.

Smoking status was based on self-reported current smoking.

The Green Care Index measures green space accessibility in residential districts. It is calculated by dividing the percentage of residential and mixed-use land within green space catchment areas (250 m–1 km radius) by population density, with a factor of 10 used to adjust the weighting. The maximum achievable score is 40 points. This index reflects green space availability relative to district characteristics.

Continuous variables were inspected for distributional shape using histograms and Q–Q plots and tested with the Shapiro–Wilk test for normality and the Levene test for homogeneity of variances. Variables with normal distributions are presented as mean ± SD and compared across the four SES groups using one-way ANOVA. Non-normally distributed variables are reported as median [IQR] and compared using the Kruskal–Wallis test. Categorical variables are presented as n (%) and compared using Pearson’s Chi-square test.

Multiple imputation was performed to handle missing data, assuming missing at random conditional on observed data, mainly due to operational factors. Predictive mean matching was used for skewed continuous variables, binary logistic for dichotomous, and ordinal logistic for ordered variables, with biologically plausible range constraints. Outcomes and follow-up times were not imputed. Analyses were pooled according to Rubin’s rules.

The primary end point was MACCE (MI, stroke, or all-cause mortality) within 60 months. Cox proportional hazards models were fitted to estimate hazard ratios (HR; 95% confidence interval [CI]) for the association with SES (4 groups), adjusting for age, sex, smoking, diabetes, hypertension, BMI, family history of CAD, known CAD, LVEF, preclinical resuscitation, and post-PCI TIMI flow. Proportional hazards assumptions were assessed. The global likelihood-ratio and visual diagnostics supported overall goodness-of-fit, and no meaningful nonlinearity of continuous covariates was observed. The analysis was illustrated using an adjusted cumulative hazards curve. To assess coefficient stability, we applied nonparametric bootstrap resampling (1,000 samples; percentile 95% CIs). As a sensitivity analysis, we performed a 90-day landmark Cox model to address early nonproportional hazards; the SES gradient remained materially unchanged, confirming the robustness of the association.

The secondary end point: A multivariable linear regression model was applied to examine whether the SES group was associated with age at first MI. The model adjusted for key clinical covariates, including sex, smoking, diabetes mellitus, arterial hypertension, obesity class III (BMI ≥40 kg/m ²), family history of CAD, and known CAD. Regression coefficients (β) and 95% confidence intervals were reported. For the linear model, residual plots and Q–Q plots indicated adequate fit. Model fit was acceptable for both the linear and Cox regression analyses.

The covariate set was defined a priori based on an extensive literature review to select the key clinical covariates. A p-value <0.05 was considered statistically significant. Statistics were performed using SPSS Statistics Version 29.

The study included a total of 2,807 patients with STEMI. Patients were stratified into four groups based on their SES: high SES (n = 552, 19.7%), intermediate SES (n = 1,551, 55.2%), low SES (n = 497, 17.7%), and very low SES (n = 207, 7.4%). Lower SES was associated with a higher rate of modifiable cardiovascular risk factors. BMI increased significantly in lower SES groups (G1: 26.1 ± 4.8 vs G4: 28.1 ± 6.8, p < 0.001 for trend) as well as the rates of diabetes mellitus (G1: 19.9% vs G4: 31.9%, p < 0.001 for trend) and smoking (G1: 33% vs G4: 54.4%, p < 0.001 for trend). No significant associations were observed in nonmodifiable cardiovascular risk factors (sex, family history of CAD, and history of CAD, all p ≥ 0.25 for trend).

Regarding lipid profiles, high-density-lipoprotein cholesterol cholesterol levels were significantly lower in very low SES patients (45 ± 10 mg/dL) compared to high SES patients (49 ± 14 mg/dL, p < 0.001 for trend). No significant differences for total cholesterol or low-density-lipoprotein cholesterol cholesterol levels were observed (all p ≥ 0.69 for trend).

Environmental exposure, assessed using the green care index, showed that patients from lower SES groups had a significantly decreased exposure to residential green spaces (p < 0.001 for trend).

The baseline characteristics of patients stratified by SES are detailed in Table 1 .

Lower SES was associated with 5.5 years younger age (G1: 68.8 ± 13.3 vs G4: 63.3 ± 12.4 years, p < 0.001 for trend). After excluding patients with known atherosclerotic disease before the index MI, the age difference increased (67.7 ± 12.4 vs 61.8 ± 13.2 years, p < 0.001 for trend). A multivariable analysis was performed to explore associations between SES classes and age of MI-onset. SES exhibits a significant and independent association with an earlier MI age even after adjustment for several classical risk factors ( Table 2 ). The multivariable analysis showed that decreasing SES levels are associated with younger age at MI presentation. Compared to patients with high SES (reference group), intermediate SES was associated with a younger age at MI onset (β = −2.63; 95% CI −3.71 to −1.56; p < 0.001), low SES (β = −3.19; 95% CI −4.54 to −1.84; p < 0.001) and very low SES (β = −3.28; 95% CI −5.05 to −1.50; p < 0.001) was associated with even earlier MI presentation.

Table 2

Linear regression analysis of socioeconomic status in association with age at myocardial infarction onset

Socioeconomic Status	Regression coefficient (β)	Standard error	95 % CI	p-value
High (Reference)	—	—	—	—
Intermediate	−2.63	0.55	−3.71, −1.56	<0.001
Low	−3.19	0.69	−4.54, −1.84	<0.001
Very Low	−3.28	0.91	−5.05, −1.50	<0.001

Only gold members can continue reading. Log In or Register to continue

Share this:

• X
• Facebook

Related

Related posts:

1. DYNAMX Bioadaptor Implantation for Treatment of Complex Coronary Artery Disease: The DYNAMITE Study
2. The Long-Term Clinical Progression of Genotype-Positive/Phenotype-Negative Patients With Hypertrophic Cardiomyopathy
3. Oversizing Transcatheter Aortic Valves: Are We Getting It Right?
4. Hidden Conduction, Overt Risk: A Case of Wolff–Parkinson–White Syndrome Unmasked

Stay updated, free articles. Join our Telegram channel

Join